Objective: To evaluate the ability of smartwatch-based measures of tremor, bradykinesia and gait to capture Parkinson’s disease (PD) progression.

Background: Wearable sensors enable remote, objective, and frequent monitoring of physical activity in people with Parkinson’s disease [1]. We have previously developed and validated a smartwatch-based active assessment, Parkinson’s Disease Virtual Motor Exam (PD-VME), for remote and self-guided assessment of different motor symptoms of bradykinesia, tremor, and gait impairment in people with PD (PwPD) [2].

Method: We included two years of at-home collected PD-VME data from the Personalized Parkinson Project (PPP) via the Verily Study Watch in 262 PwPD [3]. A PD-VME consists of eight tasks designed to assess different motor symptoms. Participants were instructed to perform the PD-VME on a weekly basis. During execution of PD-VME tasks, tri-axial accelerometer and gyroscope data were collected; the sensor data collected during off-states of symptomatic medications were used to derive digital measures of bradykinesia, tremor and gait. The following framework was used to determine whether a digital measure can track progression in motor symptoms: 1) In PwPD, the measure value changes over time with statistical significance, and the direction of change aligns with clinical expectations of disease progression (e.g., more impaired motor symptoms with progression). 2) The measure shows good month-to-month test-retest reliability (i.e., intraclass correlation coefficient (ICC) >0.75 for two consecutive months).

Results: Eight PD-VME measures showed a statistically significant change over 1 year (absolute Cohen’s D for within-subject changes ranging from 0.077 to 0.566; six measures showed greater within-subject Cohen’s D than 0.377 from MDS-UPDRS Part 3; multiplicity-adjusted p<0.05). Six of these measures showed good test-retest reliability (ICC ranging from 0.759 to 0.960). Most of these measures were derived from bradykinesia and gait-related tasks.

Conclusion: The results suggest that the PD-VME digital measures via Verily Study Watch are reliable and sensitive to capture symptom changes likely due to disease progression in PwPD. Further research to compare the PD-VME measures between people with and without PD is needed to confirm whether the change is due to disease progression.

References: [1] Kowahl et al. JMIR Hum Factors. 2023
[2] Burq et al. NPJ Digit Med. 2022
[3] Bloem et al. BMC Neurol. 2019

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/wearable-based-virtual-motor-exam-enables-remote-tracking-of-motor-symptom-progression-in-early-stage-parkinsons-disease/